如何使用 vtd-xml 解析 xml 而无需硬编码 attr 名称？

Question

这是我的示例 XML 文件，实际大小超过 2gb。使用 vtd-xml 我已经实现了这么多：

当前代码：

https://gist.github.com/shadow-fox/21d1d4f30cbed0909f403c3ac0e1fa4d

public void reader() throws IOException, ParseException, NavException, XPathParseExceptionHuge, NavExceptionHuge,
            XPathEvalExceptionHuge {
        VTDGenHuge vg = new VTDGenHuge();
    if (vg.parseFile("sku_extract_main.xml",true,VTDGenHuge.MEM_MAPPED)) {
        VTDNavHuge vnh = vg.getNav();
        AutoPilotHuge aph = new AutoPilotHuge(vnh);
        aph.selectElementNS("*", "*");
        int i = 0;
        while (aph.iterate()) {
            int t = vnh.getText();
            if (t != -1) {
                System.out.println(vnh.toString(vnh.getCurrentIndex()) + "|||" + vnh.toNormalizedString(t));
                i++;
            }
        }
    }
}

当前结果：

PVAL|||298374234
PVAL|||1231
PVAL|||brown
PVAL|||medium
PVAL|||7
PVAL|||solid
PVAL|||brown

我想要什么：

Sku_ID|||298374234
LotNum|||1231
COLOR|||brown
WIDTH|||medium
SIZE|||7
Pattern|||solid
Color Family|||brown

示例 xml：

<?xml version="1.0" encoding="UTF-8" ?>
<RECORDS>
  <RECORD>
    <PROP NAME="Sku_ID">
      <PVAL>298374234</PVAL>
    </PROP>
    <PROP NAME="LotNum">
      <PVAL>1231</PVAL>
    </PROP>
    <PROP NAME="COLOR">
      <PVAL>brown</PVAL>
    </PROP>
    <PROP NAME="WIDTH">
      <PVAL>medium</PVAL>
    </PROP>
    <PROP NAME="SIZE">
      <PVAL>7</PVAL>
    </PROP>
    <PROP NAME="Pattern">
      <PVAL>solid</PVAL>
    </PROP>
    <PROP NAME="Color Family">
      <PVAL>brown</PVAL>
    </PROP>
  </RECORD>
</RECORDS>

而且我不想硬编码attr 名称。我想在访问它们时取回它们。我该怎么做？

Answer 1

下面是我编辑的代码以打印出 attr 名称和值...它是基于 xpath...

public static void main(String s[]) throws Exception{
     VTDGenHuge vg = new VTDGenHuge();
        if (vg.parseFile("d:\\xml\\sku_extract_main.xml",true,VTDGenHuge.MEM_MAPPED)) {
            VTDNavHuge vnh = vg.getNav();
            AutoPilotHuge aph = new AutoPilotHuge(vnh);
            AutoPilotHuge aph2 = new AutoPilotHuge(vnh);
            aph.selectElementNS("*", "*");
            aph2.selectXPath("@*");
            int i = 0;
            while (aph.iterate()) {
                System.out.println(vnh.toString(vnh.getCurrentIndex()));
                int t = vnh.getText();
                if (t != -1) {
                    System.out.println(vnh.toString(vnh.getCurrentIndex()) + "|||" + vnh.toNormalizedString(t));
                    i++;
                }

// below is the my addition
// it basically evaluates the attribute axis
// push pop ensure that the node iteration of the outer while loop
// is consistent
// resetXPath is key here, without it, xpath will not work except for the
// first node returned by aph.iterate()
                vnh.push();

                while((i=aph2.evalXPath())!=-1){
                    System.out.println(" attr name "+vnh.toString(i));
                    System.out.println("attr val   "+vnh.toString(i+1));
                }
                aph2.resetXPath();
                vnh.pop();
            }
}